Tribological behavior of diamond-reinforced Fe-Co composite by - - PDF document

18TH INTERNATIONAL CONFERENCE ON COMPOSITE MATERIALS

• 1. Introduction

Diamond has the highest thermal conductivity and low coefficient of friction, and shows superior mechanical properties such as bulk modulus, tensile stress and hardness. High hardness and wear resistance result in good surface finish and long tool life [1]. Therefore, diamond tools are widely used for difficult-to-cut materials like Al-Si alloys, fiber reinforced composite or stone, concrete. The matrix used in the most diamond tool, is a cobalt alloy since it combines with diamond grits good chemical compatibility at the processing temperature. However, the price of Co is subjected to great variations [2]. This is the main reason behind efforts to replace Co with other metals such as Co-bronze, Fe-bronze, Tin-Co, Fe-Co, Fe-Cu, Fe-Co-Cu materials [3, 4]. The properties of diamond abrasive and the adhesion between the diamond grits and bond determine properties of diamond tools [5]. This also depends upon composition of metal powders, their sizes and distributions, processing temperature and times [6]. The aim of the present study is thus, to investigate the wear behavior of diamond impregnated composite by changing the matrix composition based on Fe/Co ratio and B4C powder addition.

• 2. Experimental study

Fe-Co-Bronze based diamond with and without B4C reinforcement composites were produced by hot pressing method. Diamonds were used as reinforcement in the production of metal matrix composite while Co metal was a binder with a good wet ability. Fe is also used as a filling material because of its lower-cost. The concentration of diamond grits in the composite was kept at about 20%. Fe-(15-35-55wt. %Co)-bronze 10wt. %- diamond 20wt. %. A Co, Fe powder and diamond grits with an average particles size of 400 mesh, 85/15 mesh and -40/+50 mesh powders were selected as the starting material. Hereafter, these are denoted by C1 (Fe-15Co-10bronze-20diamond), C3 (Fe-35Co-10bronze-20diamond), and C5 (Fe-55Co- 10bronze-20diamond). In the production of composites, Fe-Co-bronze- diamond- was blended in T2 turbula mixture at about 45 min. The mixture per composition was put in carbon moulds 24X10X10 mm dimension were cold compacted, and then hot zone was evacuated to remove the air from the chamber atmosphere, and sintered by hot pressing of Fritsch DSP 510 type machine under nitrogen atmosphere. The sintering temperature was about 800 oC under a compression

• f 25 MPa. Total sintering time was about 15 min.

The wear parameters were abrasive size (A), material hardness (B), load (C) and sliding distance (D) and their levels indicated in Table 1. An L18

• rthogonal array was chosen.

Tribological behavior of diamond-reinforced Fe-Co composite by Taguchi method

• K. Emre Öksüz 1, M. Şimşir 1, Y. Şahin 2*

1Cumhuriyet University , Metallurgical and Materials Engineering Department, 58140,

Sivas/Turkey, 2 Gazi University, Manufacturing Engineering Department, Faculty of Technology, 06500 Ankara/Turkey * Corresponding author(ysahin@gazi.edu.tr)

• Abstract. Tribological behavior of diamond-reinforced Fe-Co composites was studied. SN ratio and

ANOVA were employed to investigate the influence of parameters like, abrasive size, load, distance and

• hardness. The load factor was the major parameter, followed by abrasive size. The percent contributions

are abrasive size (32.17 %), hardness (11.30 %). Keywords: Fe-Co matrix; Diamond tool; Powder metallurgy; Tribological behavior .

18th International Conference on composite materials

Table 1. Control factors and their levels for diamond reinforced MMCs.

Levels Abras. size, µm Hardness , BHN Applied load, N Sliding dist., m 1 22 µm 76.6 5 24 2 68 µm 79 15 48 3

• 84

25 72

• 3. Results and discussion

3.1. Main effects plot Fig.1a indicates the main effects plots of the weight loss of composites for the means. The average values of weight loss for each parameter at levels 1, 2 and 3 were calculated and were shown in Fig.1. Optimal testing conditions of these control factors could be determined from this graph. The best weight loss value was at the lower mean value in the response graph. It was clear from Fig.1 that the weight loss was minimum at the 1st level of parameter A, 3rd level of parameter B, 1st level of parameter C and lastly 1st level of parameter D. It was evident that the load (C) had the greatest effect

• n the optimal testing conditions. This might be

because SiC abrasive did not have an ability to dig

• ut of diamond powders in its place to remove it. On

the other hand, the level of a factor with highest S/N ratio is the optimum level, as shown in Fig.1b. Hence, the optimum levels of the factors are A1, B3, C1 and D2. Table 2 shows experimental lay out and results with calculated S/N ratios for weight loss of the

• composites. The control factor with the strongest

influence is detected by difference value. It can be seen in Table 2 that the strongest influence is exerted by C and A, respectively. The weight loss obviously increases as load increases from 5N to 25N, followed by the abrasive particle size from 400 mesh (≈22µm) to 180 mesh ((≈68µm). In addition, weight loss increases slightly with the B and D factor. These effects, however, much lower compare to those of C and A factor. The results are good agreement with Mondal et al. [7], Prasad et al. [8] and Basavarajappa et al. [9]. However, previous works by Fernandez et al. [10], Sahin [11] indicated that abrasive size was the more effective on wear rate of the samples. Basavarajappa et al. [9] also found that the sliding distance was the most effective parameter for Al/SiCp and Al/Gr matrix composites, which is followed by the load. The least weight loss is recorded during 7th test run when the process parameters were at first level of A, 3rd level

• f B, first level of C and D factor. It is followed by

8th, 4th and 1st run.

68 22

.0 2 .0 1 5 .0 1

84 .0 79.0 76.6 2 5 1 5 5

.0 2 .0 1 5 .0 1

72 48 24 A

Mean of Weight Loss (g)

B C D

M a in E f fects P lo t for M ea n s

a) Mean

6 8 2 2 4 2 4 3 8 3 6 3 4 8 4 .0 7 9 .0 7 6 .6 2 5 1 5 5 4 2 4 3 8 3 6 3 4 7 2 4 8 2 4

A

Mean of SN ratios (dB)

B C D

M a in E ffec ts P lot fo r S N r a tios

b) S/N ratio Fig.1. Effects of main parameters on the weight loss of diamond-reinforced MMCs. a) Abrasive size, b) Hardness, c) Load, d) Sliding distance.

18th International Conference on composite materials

The wear performance of the tested samples was estimated from Eq.(1) using the optimal testing factors. (1) The average performance of S/N ratios was found to be about 49.376dB. It could be noted that the

• ptimum condition for the "smaller is better" quality

characteristic was A1 B3 C1 D2. Optimal settings of control factors for the tested samples were: load (5N), abrasive size (22 µm), hardness (84HB) and sliding distance (48m). Table 2. Experimental lay out and results with calculated S/N ratios for weight loss of MMCs.

Exp. no. Abr. size, µm Mater. hardn., HB (B) Load ,N (C) Slid. dist., m Weight loss (g) S/N ratio (dB) (A) (D) 1 22 76.5 5 24 0.004 47.958 2 22 76.5 15 48 0.0082 41.723 3 22 76.5 25 72 0.020 33.979 4 22 79 5 24 0.0068 43.349 5 22 79 15 48 0.0077 42.270 6 22 79 25 72 0.0145 36.772 7 22 84 5 48 0.0027 51.372 8 22 84 15 72 0.0065 43.741 9 68 84 25 24 0.0132 37.588 10 68 76.5 5 72 0.0106 39.493 11 68 76.5 15 24 0.030 30.457 12 68 76.5 25 48 0.0333 29.551 13 68 79 5 48 0.0091 40.819 14 68 79 15 72 0.0169 35.442 15 68 79 25 24 0.0184 34.703 16 68 84 5 72 0.0154 36.249 17 68 84 15 24 0.014 37.077 18 68 84 25 48 0.0172 35.239

The response table of the weight loss is presented in Table 3. It indicates that the mean S/N ratios at each level of control factor and how it is changed when settings of each control factor are changed from level 1 to level 2. The influence of interactions between control factors is ignored. The control factor with the strongest influence is detected by difference value. The higher the difference, the more influential is the control factor. It can be seen that the strongest influence is expected by factor C and factor A, respectively.

Table 3. The response table for signal to noise ratios of MMCs. Weight loss of MMCs (dB) Level A B C D 1 42.08 37.19 43.21 38.52 2 35.45 38.89 38.45 40.17 3

• 40.22

34.65 37.61 Differ 6.63 3.03 8.56 2.56 Rank 2 3 1 4

3.2. Confirmation test The confirmation tests are performed by selecting the set of parameters. Table 4 shows the confirmation test results and a comparison of the predicted weight loss using the optimal testing

• parameters. The values obtained experimentally, the

values obtained from the optimal settings were compared, and some differences were found between experimental and theoretical values. A good agreement between the predicted and actual weight loss was observed. It can be seen in Table 4 that the predicted error of S/N ratio was in the reasonable

• limit. The confirmation tests showed that error

associated with abrasive wear of the composite was about 3, 35%. The deviation is found to be less from the predicted values.

Table 4. The confirmation test results.

Performance measure Results Levels (A,B,C,D) 1,3,1,2 S/N observed, dB 46.020 S/N predicted, dB 49.370 Prediction of /N ratio, dB 3.35

3.3. Analysis of Variance (ANOVA) Table 5 shows the mean results of analysis of variance (ANOVA) for the wear of the samples. This analysis is performed with a level of significance of 5%. The last column of the table shows the contribution % (p) of each variable in the total variation indicating the influence degree on the abrasive wear of sample. If the "Test F" value is greater than the F (5%) column value, then the assigned variable is statistically significant. It can be

• bserved from the ANOVA table that the applied

load (p=33.04%) and the abrasive size (p=32.17%) has great influence on the wear. However, hardness

18th International Conference on composite materials

(11.30%) and sliding distance (0.52%) does not have a significant effect on the abrasive wear for both physical and statistical point of view. The residual error associated in the ANOVA table is approximately about 22.97%. This shows clearly as the applied load and abrasive size increases the wear rate also increases in the most of the cases. From this table, it indicates that applied load has the major contribution for the weight loss compare to other parameters.

Table 5. Results of the mean-ANOVA for diamond reinforced Fe-Co matrix composite. Sour . of var. D O F SS Variance Test F F P,% A 1 0.00037 0.00037 14.4 4.96 32.1 B 2 0.00013 0.00006 2.70 4.10 11.3 C 2 0.00038 0.00019 7.59 4.10 33.0 D 2 0.00000 6 0.000003 0.12 4.10 0.5 Res. error 10 0.00025 0.000025 22.9 Tot. 17 0.00115 100

In order to study the significance of the parameters in affecting the quality characteristic of interest for wear data, ANOVA in terms of S/N ratio was

• performed. The S/N ANOVA version is given in

Table 6. It is clear from this table that the parameter C significantly affects the wear of the composite, which is followed by the parameter A, again.

Table 6. Results of the S/N-ANOVA for diamond reinforced Fe-Co matrix composite. So. D F SS Varian. Test F F P,% A 1 197.83 197.83 24.82 4.96 36.2 B 2 27.61 13.80 1.73 4.10 5.05 C 2 220.7 220.7 13.85 4.10 40.4 D 2 20.17 20.17 1.27 4.10 3.69 Res err. 1 79.69 7.969 14.6 Tot. 1 7 546.01 100

It can be observed from the ANOVA table that the applied load (p=40.42%) and the abrasive size (p=36.23%) have great influence on the wear, but hardness (p=5.05%) and sliding distance (p=3.69%) do not have a significant effect on the abrasive wear. Optimal setting of control parameters for the abrasive wear of MMCs against SiC abrasives are: load (5N), abrasive size (22 µm), hardness (84HB) and sliding distance (48m). The percent contribution

• f parameters based on mean weight loss in

affecting variation for abrasive wear of MMCs are: abrasive size (32.17 %), hardness (11.30 %), load %33.04 %), and sliding distance (0.52 %). The percent contribution of parameter in terms of S/N ratio is also confirmed by the ANOVA.

References [1] Y. Şahin. “Metal cutting principles”. 2nd edition, Seçkin Publication, Ankara, 2001. [2] I. E. Clark and B. J. Kamphuis. “Recent advances in prealloyed powders for diamond tooling”. Proceedings of European Conf. on Hard Mater. Diam. Tools by EPMA, Lausanne, Switzerland, pp.35-42, 2002. [3] L. J. Oliveira, G. S. Bobrovnitchi and M. Filgueria. “Proceeding and characterization of impregnated diamond cutting tools using a ferrous metal matrix”. Int.J.Refrac.Met.Hard Mater.25, pp.328-335, 2007. [4] Z. Nitkiewicz and M. Swierzy. “Tin influence on diamond-metal matrix hot pressed tools for stone cutting”. J.Mater. Process.Tech.175, pp.306-315, 2006. [5] Y. H. Wang, J. B. Zang, M. Z. Wang, Y. Guan and Y.

• Z. Zheng. “Properties and applications of Ti-coated

diamond grits”. J.Mater.Process.Tech.120, pp.369-372, 2002. [6] Y. S. Ko, T. Tsuruni and O. Fukunage. “High pressure sintering of diamond-SiC composite”. J.Mater.Sci.36, pp.469-475, 2001. [7] D. P. Mondal, S. Das, A. K. Jha and A. H.

• Yegneswaran. “Abrasive wear of Al alloy-Al2O3 particle

composite: A study of the combined effect of load and size of abrasive”. Wear 223, pp.131-138, 1998. [8] B. K. Prasad, S. Das, A. K. Jha, O. P. Modi, R. Dasgupta, A. H. Yegnaswaran. “Wear of Al2O3 particle composite”. Composites 28A, pp.301-308, 1997. [9] S. Basavarajappa, K.V. Arun and J. Davim. “Effect of fillers on dry sliding wear behavior of polymer matrix composites-A Taguchi approach”. J.Miner.Mater.Caharact.Eng.8, pp.375-391, 2009. [10] J. E. Fernandez, M. R. Fernandez, R. V. Diaz and R.

• T. Navarro. “Abrasive wear analysis using factorial

design”. Wear 255, pp.38-43, 2003. [11] Y. Sahin. “The prediction of wear resistance model for the metal matrix composites”. Wear 258, pp.1717- 1722, 2005. [12] S.Basavarajappa, G.Chandramohan and J.Paulo

• Davim. “Application of Taguchi techniques to study dry

sliding wear behaviour of metal matrix composites”.

• Mater. Des.28, pp 1393-1398, 2007.

18th International Conference on composite materials